MLUtils.jl
MLUtils.jl copied to clipboard
JuliaML
Utilities and abstractions for Machine Learning tasks
[FLoops](https://github.com/JuliaFolds/FLoops.jl) is no longer mantained and it has not been forked yet by the JuliaFolds2 organization (see https://github.com/JuliaFolds2/Transducers.jl/issues/31) What are the available alternatives for our DataLoader implementation? Would the simple...
Now that MLUtils.jl has feature parity with `DataLoaders.DataLoader`, it's time to deprecate DataLoaders.jl. We should: - [ ] add a guide for users transitioning that explains the differences between `DataLoaders.DataLoader`...
I am increasingly relying on the `getobs`/`nobs` interface in quite low-level packages I am working on. It's nice to be able to work generically with tables and arrays. But I...
Is `DataLoader` type-unstable? Seems so. Does it have to be type-unstable? ``` using Flux function test_data_loader(data_loader) for (x,y) in data_loader continue end end X = rand(Float32, 2, 10) Y =...
It's surprising that `splitobs` and `DataLoader` make views, when they mention only `getobs` in their docstrings, which does not: ```julia help?> splitobs splitobs(data; at, shuffle=false) -> Tuple Split the data...
The keyword `at` is described as a proportion, but secretly has quite different behaviour when it's an integer. IMO it would be clearest if these had distinct names, but if...
Hi, have you thought about porting some Time Series utility functions? Such as `kfold` for time series? https://alan-turing-institute.github.io/MLJ.jl/stable/evaluating_model_performance/#MLJBase.TimeSeriesCV ```julia julia> MLJBase.train_test_pairs(TimeSeriesCV(nfolds=3), 1:10) 3-element Vector{Tuple{UnitRange{Int64}, UnitRange{Int64}}}: (1:4, 5:6) (1:6, 7:8) (1:8,...
DataLoader performance degrades through iterations when `parallel=true`. Following is a MWE illustrating the issue: ```julia using Images using StatsBase: sample, shuffle using DataAugmentation using Flux using TestImages import Base: length,...
In particular, the following does not work ```julia data1 = 1:10 data2 = 11:20 data12 = joinobs(data1, data2) data3 = 21:30 data123 = joinobs(data12, data3) ``` The error I get...
fix #156 This covers only `BatchView(...; collate=nothing)`, I'm not sure how `collate=true` and `collate=false` should interact with the buffer so I won't address that here.